Palladium-catalyzed modular biomimetic synthesis of lignans derivatives
Junlong Tang , Yuhan Zhao , Yangbin Jin , Liren Zhang , Yuanfang Wang , Wanqing Wu , Huanfeng Jiang
From 1981 to 2019, more than one-third of the small-molecule drugs approved by the FDA were derived from natural products and their derivatives [1]. Expanding natural product scaffolds to elucidate structure-activity relationships is essential for enhancing drug properties and speeding up drug discovery, which has been intriguing and sought after in organic synthesis [2–5]. Notably, lignans, plant-derived natural products with notable bioactivities, are commonly composed of phenylpropanoid (C6-C3 unit) dimers linked by the C8 carbon atom [6–9]. The diverse cyclization patterns, oxygenation states, and aromatic moiety substitutions of phenylpropanoid contribute to the structural variability of lignans, allowing for a wide range of bioactivities despite their modest size and complexity (Scheme 1A, left) [10–14]. A wide range of promising pharmacological activities, including antitumor, antiviral, anti-inflammatory, antimicrobial, antioxidant, antiallergic, antihypertensive, hepatoprotective, and sedative, have been discovered. For instance, NAGA extracted from Larrea tridentata, a traditional herb, functions as a 5-lipoxygenase and tyrosine kinase inhibitor [15]. Podophyllotoxin and its derivatives are utilized in cancer therapies [16]. Hinokinin, isolated from Pycnanthus angolensis, demonstrates significant apoptosis-inducing effects on human hepatoma cells [17]. Acutissimalignan B, derived from Phyllanthus acutissima, exhibits inhibitory properties on HIV-1 cells (Scheme 1A, right) [18].
Significant importance has been placed on the chemical synthesis of lignans due to challenges in isolating them from plants, including yield, complexity, and cost. Over the last decade, considerable advancements have been achieved in this area with various elegant strategies [19–29]. Remarkably, in 2020, Zhu [27] reported a cutting-edge visible-light photoredox-catalyzed [4 + 2] cycloaddition method for the synthesis of aryltetralin cyclic ether lignans. Lumb [28] developed a pioneering mimicking lignan biosynthesis utilizing redox-neutral photocatalyzed oxidative radical cyclization. Despite significant progress, to some extent, clever substrate designs or multiple manipulations were required, and collective synthesis of lignan skeletons was less reported. Besides, the precise synthesis of lignan through diversified editing of the phenylpropanoid unit remains underdeveloped. Structure-activity relationship studies have shown that even subtle changes in the phenylpropanoid structure can significantly impact the microtubule depolymerizing activity and therapeutic window of the scaffold of lignans [30,31]. Therefore, at the phenylpropanoid level, a modular, editable, and cost-effective lignan synthesis in a single operation is a long-standing desire yet remains unrevealed, which plays a crucial role in discovering lignan-based drugs.
Inspired by the biosynthesis of lignans (Scheme 1B, left), we envisioned developing a biomimetic synthesis strategy: constructing two phenylpropanoid molecules and coupling them at the C8 positions in situ. With the vision and our research interests in cyclization reactions involving alkenes and alkynes [32–34], a retrosynthetic analysis of dehydrodibenzylbutyrolactone, a crucial class of lignans and a key intermediate for synthesizing other lignans, revealed that a concise sequence of C(sp2)-arylation, C(sp2)-C(sp3) coupling and C(sp3)-arylation could efficiently produce the target lignan using three raw synthons: arylboronic acid, propiolic acid, and allyl halide acid (Scheme 1B right). Therefore, we envision achieving the palladium-catalyzed selective dicarbofunctionalization of 1, 6-enyne to modularize the synthesis of dehydrodibenzylbutyrolactone. Considering assembly sequence and stereoselectivity, this proposal remains a formidable challenge, as some key issues must be addressed: (1) Several well-established transformations of organopalladium species take priority over the designed sequences. Specifically, reported dicarbofunctionalization reactions have primarily involved electron-rich internal alkynes, leaving the potential applicability of electron-deficient alkynes such as alkyne esters underexplored [35–40]. Additionally, without stabilization of the directing group or the benzyl and allyl sites, the selective transmetalation and reductive elimination of alkyl palladium(Ⅱ) species with β-H remains underreported [41–46]. (2) Precise control of the Z/E selectivity of the C(sp2)-arylation bond is crucial. (3) The rapid decomposition of 1, 6-enyne esters in the presence of Pd(0) poses a significant challenge (Scheme 1C) [47,48]. Herein, we developed a palladium/phosphine ligand/benzoquinone system in which the selective oxidative dicarbofunctionalized cyclization of 1, 6-enyne was accessible. This protocol allows for the modular synthesis of dibenzylbutyrolactone-derived lignans from simple chemical feedstocks. Notably, the versatility of the three building blocks and the ease of product modification facilitate the efficient assembly and diversification of lignan libraries, enabling the streamlined synthesis of various lignan skeletons.
Considering the abovementioned challenges, we inferred that ligand control and oxidative regeneration of palladium(Ⅱ) species were crucial for achieving the envisioned domino sequence. To test our hypothesis, we explored reaction conditions with 1, 6-enyne esters 1a and phenylboronic acid 2a as model substrates. Unfortunately, initial screening did not yield the target product 3a; instead, two by-products 4 and 7 were observed, likely resulting from the protonolysis of Csp2-Pd species and the coupling between decomposed 1a and 2a (see Supporting information for more details). To our delight, we isolated 3a in 25% yield with the assistance of DPPP and benzoquinone, along with two new cyclic by-products 5 and 6 from the protonolysis and β-H elimination of Csp3-Pd species. Ultimately, we found that the combination of Pd(TFA)2 (10 mol%) as the catalyst, DPPF (15 mol%) as the ligand, and THF (0.1 mol/L) as the solvent resulted in efficient oxidative dicarbofunctionalization of 1, 6-enyne (Fig. 1). To summarize critical observations, our analysis revealed that the use of alternative palladium(Ⅱ) salts, such as PdCl2 and Pd(OAc)2, resulted in notably decreased yields, while palladium(0) precatalysts, like Pd2(dba)3, appeared to cause the decomposition of 1a, suggesting a potential link between the Lewis acidity of palladium and chemoselectivity. Variations in ligands, oxidants, solvents, and reaction temperature significantly impacted the outcome, with different choices leading to lower yields or altered chemoselectivity. The distinct behavior observed with benzoquinone (BQ) and p-toluquinone (2-MeBQ) suggested a possible coordination effect between quinones and palladium, highlighting the influence of steric effects on the catalytic cycle.
With the optimized conditions in hand, we explored the scope of this protocol (Scheme 2). Generally, experiments demonstrated compatibility with a wide range of organoboronic acids and enynes. It was observed that varying the aromatic substituents on the aryl boronic acids was well tolerated, forming products 3a-3t in moderate to good yield. A broad range of electron-donating groups were well-tolerated, such as -OMe (3e, 3r), -OBn (3f), -SMe (3g), -NHBoc (3i), morpholino group (3j), as well as electron-withdrawing groups like -OCF3 (3h), -F (3k), -CF3 (3l), -Ac (3m, 3o), -CO2Me (3n, 3s) with either para-substituents or meta-substituents. Since the polyoxy-substituted aromatic ring of lignans significantly influences bioactivities, we investigated the compatibility of polyoxy-substituted aryl boronic acids and found that several familiar multi-substitution groups among lignans were well compatible with 72%−88% yield (3u-3x). Notably, medicinally prevalent aromatic moieties other than benzene were incorporated smoothly, including naphthyl (3y), furyl (3z), thiophenyl (3aa), pyrrolyl (3ab), benzofuranyl (3ac), benzothiophenyl (3ad), indolyl (3ae) and dibenzothiophenyl (3af). Much to our delight, the alkenyl-boronic acid also matched our reaction system, providing the lignan derivatives containing a 1.3.7-triene skeleton, covering internal alkene (3ag, 3aj), terminal alkene (3ah) and cycloalkene (3ai), which would be cumbersome to obtain by other methods. The tolerance of functional groups on the propiolic acid building blocks was also examined. It was found that alkyne units containing electron-withdrawing groups at the p-position, such as -F (3ba), -Br (3bb), and -CO2Me (3bc) were well tolerated, providing products with 60%−80% yields. The cyano (3bd) moiety was also incorporated into the lignan without poisoning the reaction efficiency. The corresponding products with -tBu (3be), -Ph (3bf), and -OEt (3bg) were well constructed. Steric and electrical effects were evaluated with o, m-substituents (3bh-3bk). The decreased yield could originate from the unfavorable steric hindrance. Multi-substituted aromatic and heteroaromatic enynes (3bl-3bn) worked smoothly, offering products with 60%−73% yields. Substrates with simple alkyl substituents, such as -H, -Me, -Et, -Pr (3bo-3br) provided 54%−65% yields. This protocol was further successfully extended to the allyl halide block substituted at C1 (3bs) and C3 (3bt). Delightfully, lactam-based lignans derivatives (3bu, 3bv) could be constructed in moderate yield.
Molecular hybridization is a crucial and practical approach in drug design, combining various pharmacophore fragments to create a more potent bioactive molecule [49,50]. Inspired by the positive pharmacological effect of lignans and the diversity of building blocks, we explored the potential of our method for the modular molecular hybridization of lignans with natural products, biologically active compounds, and pharmaceutical agents (Scheme 3). Utilizing drug-derived arylboronic acids, a novel lignan skeleton incorporating two pharmacophore fragments was efficiently constructed. Notably, commercially available pharmaceuticals such as sesamol (3ca), clofibrate (3cb), fenofibrate (3cc), and indomethacin (3cd) fragments were successfully incorporated into γ-butyrolactone with 53%−72% yields. Perillyl alcohol (3ce), and l-menthol (3cf) were also demonstrated to hybridize with lignans seamlessly. Profiting from the convenient synthesis of 1, 6-enyne, pharmacodynamic moieties were successfully introduced into the non-aryl ring framework of lignans. Various drugs such as THFA (3cg), ibuprofen (3ci), roflumilast (3cj), flurbiprofen (3ck), gemfibrozil (3cl), and oxaprozin (3cm) were able to undergo formal molecule hybridization with lignans. Surprisingly, clodinafop-propargyl (3ch), an alkyl alkyne with multiple functional groups, including catalyst-poisoning pyridinyl, displayed effectiveness with moderate yields, highlighting the broad scope and good functional-group compatibility of the approach.
In addition to the high throughput screening of aromatic moiety substitutions in the phenylpropanoid unit, the collective synthesis of lignan skeletons with different cyclization patterns and oxygenation states is crucial in discovering lignan-based drugs. As such, we intended to realize the divergent synthesis of lignan skeletons to showcase the potential of our methodology in drug optimization studies (Scheme 4). With 3, 4-(methylenedioxy)phenylboronic acid as the template aryl source, (rac)-savinin (8) was obtained in a single operation with 48% yield, which, to the best of our knowledge, was the most concise process in the literature [51]. The diastereodivergent reduction of the double bond in 8 yielded (rac)-hinokinin 9 and (rac)-isohinokinin 10, respectively, with the assistance of Ni/NaBH4 and Mg/MeOH, allowing for the synthesis of unnatural diastereoselective lignans. Furthermore, the double bond of (rac)-savinin 8 underwent an oxidative Heck reaction to achieve Csp2H arylation, forming 17, a lignan skeleton with triaryl groups. Subsequent treatment with various reagents led to the presence of (rac)-cubebin 11, (rac)-dehydroxycubebin 12, (rac)-dihydrocubebin 13, and (rac)-dehydrocubebin 14 through the selective reduction of the ester moiety in (rac)-hinokinin. Dibenzocyclooctadiene lignans 15 and 16 could be rapidly constructed by oxidative coupling of aromatic rings.
To elucidate the mechanism and selectivity-origin of the reaction, a series of experiments were conducted (Scheme 5). Initially, in the absence of DPPF, only non-cyclic product 4 was obtained (Scheme 5A). Further experiments revealed that 4 was not an intermediate to the target compound 3 (Scheme 5B). Methylphenylpropynoate 18, allyl acetate 19, and phenylboronic acid 1b were subjected to both "control conditions" and "standard conditions." Results indicated that in the absence of DPPF, the migration and insertion of the triple bond into the Ar-Pd(Ⅱ) species was difficult, providing a few products 20. However, with DPPF, a coupling product between the Csp2-Pd species and alkene 19 was observed, yielding 21 in 35% (Scheme 5C). Consequently, DPPF could accelerate the migration and insertion of double bonds into the Csp2-Pd species and partially inhibit the protonolysis of the Csp2-Pd species. Subsequently, the impact of the ligand on the reduction and elimination of the Csp3-Pd-Ar species was explored (Scheme 5D). In the absence of DPPF, only the oxidative-Heck-type product 23 was detected in 22% yields, which indicated the Csp3-Pd species would rapidly undergo β-H elimination, making selective transmetalation and reductive elimination unfeasible. Under standard conditions, diarylation product 24 was obtained with a 53% yield, highlighting the importance of DPPF in the reductive elimination of the Csp3-Pd-Ar species.
In summary, at the phenylpropanoid level, a general, step-economy, and value-added methodology for the modular synthesis of lignans involving palladium-catalyzed dicarbofunctionalized cyclization of 1, 6-enyne was developed. The diversity and the high editability of three raw synthons generated powerful access to the rapid construction of lignan libraries and the collective syntheses of lignan skeletons. The high efficiency of the molecule hybridization of lignans with drugs also highlighted the modular synthesis. Combining phosphine ligands with quinones played a crucial role in switching chemoselectivity to ensure the feasibility of the design program. Further application of structurally novel lignan derivatives synthesized by the methodology is ongoing in our laboratory.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Junlong Tang: Conceptualization. Yuhan Zhao: Formal analysis. Yangbin Jin: Methodology. Liren Zhang: Software. Yuanfang Wang: Software. Wanqing Wu: Writing – review & editing. Huanfeng Jiang: Writing – review & editing.
Financial support was provided by the State Key Laboratory of Pulp and Paper Engineering (No. 2022PY01), the National Natural Science Foundation of China (Nos. 22231002 and 21871095), and the Key-Area Research and Development Program of Guangdong Province (No. 2020B010188001)
Supplementary material associated with this article can be found, in the online version, at doi:
D.J. Newman, G.M. Cragg, J. Nat. Prod. 83 (2020) 770–803. doi: 10.1021/acs.jnatprod.9b01285
T. Rodrigues, D. Reker, P. Schneider, G. Schneider, Nat. Chem. 8 (2016) 531–541. doi: 10.1038/nchem.2479
I.B. Seiple, Z. Zhang, P. Jakubec, et al., Nature 533 (2016) 338–345. doi: 10.1038/nature17967
M. Chen, J. Liu, P. Duan, M. Li, W. Liu, Natl. Sci. Rev. 4 (2017) 553–575. doi: 10.1093/nsr/nww045
L. Li, Z. Chen, X. Zhang, Y. Jia, Chem. Rev. 118 (2018) 3752–3832. doi: 10.1021/acs.chemrev.7b00653
E. Fuss, Phytochem. Rev. 2 (2003) 307–320.
J.Y. Pan, S.L. Chen, M.H. Yang, et al., Nat. Prod. Rep. 26 (2009) 1251–1352. doi: 10.1039/b910940d
W. Lau, E.S. Sattely, Science 349 (2015) 1224–1227. doi: 10.1126/science.aac7202
R.B. Teponno, S. Kusari, M. Spiteller, Nat. Prod. Rep. 33 (2016) 1044–1092.
Q. Cui, R. Du, M. Liu, L. Rong, Molecules 25 (2020) 183–200. doi: 10.3390/molecules25010183
N. Dong, H. Lin, J. Integr. Plant Biol. 63 (2021) 180–209. doi: 10.1111/jipb.13054
A. Pelter, vol 20. Springer, Boston MA.
M. Saleem, H.J. Kim, M.S. Ali, Y. Lee, Nat. Prod. Rep. 22 (2005) 696–716. doi: 10.1039/b514045p
S. Xiao, M. Gao, Z. Sun, et al., Eur. J. Med. Chem. 208 (2020) 112830–112854.
W.S. Lee, Y.I. Baek, J.R. Kim, et al., Med. Chem. Lett. 14 (2004) 5623–5628.
X. Yu, Z. Che, H. Xu, Chem. Eur. J. 23 (2017) 4467–4526. doi: 10.1002/chem.201602472
T.A. Mansoor, R.M. Ramalho, C.M.P. Rodrigues, et al., Phytother. Res. 26 (2011) 692–696.
P. Tuchinda, J. Kornsakulkarn, M. Pohmakotr, et al., J. Nat. Prod. 71 (2008) 655–663. doi: 10.1021/np7007347
P.S. Baran, M.P. DeMartino, Angew. Chem. Int. Ed. 45 (2006) 7083–7086. doi: 10.1002/anie.200603024
H. Weinstabl, M. Suhartono, Z. Qureshi, M. Lautens, Angew. Chem. Int. Ed. 52 (2013) 5305–5308. doi: 10.1002/anie.201302327
C.P. Ting, T.J. Maimone, Angew. Chem. Int. Ed. 53 (2014) 3115–3119. doi: 10.1002/anie.201311112
W. Fu, M. Nie, A. Wang, Z. Cao, W. Tang, Angew. Chem. Int. Ed. 54 (2015) 2520–2524. doi: 10.1002/anie.201410700
E.R. Welin, A.A. Warkentin, J.C. Conrad, D.W.C. MacMillan, Angew. Chem. Int. Ed. 54 (2015) 9668–9672. doi: 10.1002/anie.201503789
G. Liao, Q. Yao, J. Zhang, et al., Angew. Chem. Int. Ed. 57 (2018) 3661–3665. doi: 10.1002/anie.201713106
J. Chern, C. Lu, Z. Fang, et al., Angew. Chem. Int. Ed. 57 (2018) 7040–7045. doi: 10.1002/anie.201801618
M. Lazzarotto, L. Hammerer, M. Hetmann, et al., Angew. Chem. Int. Ed. 58 (2019) 8226–8230. doi: 10.1002/anie.201900926
J.C. Xiang, Q. Wang, J. Zhu, Angew. Chem. Int. Ed. 59 (2020) 21195–21202. doi: 10.1002/anie.202007548
Z. Huang, J.P. Lumb, Nat. Chem. 13 (2020) 24–32.
M. Xu, M. Hou, H. He, S. Gao, Angew. Chem. Int. Ed. 60 (2021) 16655–16660. doi: 10.1002/anie.202105395
N.S. Plaha, S. Awasthi, A. Sharma, N. Kaushik, 3 Biotech 12 (2022) 255–273.
S. Shen, Y. Tong, Y. Luo, L. Huang, W. Gao, Nat. Prod. Rep. 39 (2022) 1856–1875. doi: 10.1039/d2np00028h
W. Yu, B. Zhu, F. Shi, et al., Angew. Chem. Int. Ed. 60 (2021) 1313–1322. doi: 10.1002/anie.202010889
C. Liu, X. Tan, L. Zhan, et al., Angew. Chem. Int. Ed. 61 (2022) e202215020.
C. Liu, J. Wu, X. Tan, et al., ACS Catal. 13 (2023) 11339–11344. doi: 10.1021/acscatal.3c02864
T. Satoh, S. Ogino, M. Miura, M. Nomura, Angew. Chem. Int. Ed. 43 (2004) 5063–5065.
Y. Li, X. Liu, H. Jiang, Z. Feng, Angew. Chem. Int. Ed. 49 (2010) 3338–3341. doi: 10.1002/anie.201000003
M. Jiang, T. Jiang, J.E. Bäckvall, Org. Lett. 14 (2012) 3538–3541. doi: 10.1021/ol301551x
M. Jiang, J.E. Bäckvall, Chem. Eur. J. 19 (2013) 6571–6575. doi: 10.1002/chem.201204555
S. Dutta, S. Yang, R. Vanjari, et al., Angew. Chem. Int. Ed. 59 (2020) 10785–10790. doi: 10.1002/anie.201915522
A. Whyte, A. Torelli, B. Mirabi, et al., J. Am. Chem. Soc. 142 (2020) 9510–9517. doi: 10.1021/jacs.0c03246
S.G. Newman, M. Lautens, J. Am. Chem. Soc. 133 (2011) 1778–1780. doi: 10.1021/ja110377q
K. Wang, Z. Ding, Z. Zhou, W. Kong, J. Am. Chem. Soc. 140 (2018) 12364–12368. doi: 10.1021/jacs.8b08190
E. Yamamoto, M.J. Hilton, M. Orlandi, et al., J. Am. Chem. Soc. 138 (2016) 15877–15880. doi: 10.1021/jacs.6b11367
Y. Zhang, G. Chen, D. Zhao, Chem. Sci. 10 (2019) 7952–7957. doi: 10.1039/c9sc02182e
Y. Xi, W. Huang, C. Wang, et al., J. Am. Chem. Soc. 144 (2022) 8389–8398. doi: 10.1021/jacs.2c03411
T. Xia, Y. Xi, H. Ding, et al., Chem. Commun. 58 (2022) 9282–9285. doi: 10.1039/d2cc03202c
Z. Jiang, S.L. Niu, Q. Zeng, et al., Angew. Chem. Int. Ed. 60 (2021) 297–303. doi: 10.1002/anie.202008886
K.N. Flesch, A.Q. Cusumano, P.J. Chen, et al., J. Am. Chem. Soc. 145 (2023) 11301–11310. doi: 10.1021/jacs.3c02104
G. Wess, M. Urmann, B. Sickenberger, Angew. Chem. 113 (2001) 3443–3453.
C. Viegas-Junior, A. Danuello, V. da Silva Bolzani, E.J. Barreiro, C. Fraga, Curr. Med. Chem. 14 (2007) 1829–1852.
T. Golakoti, H.K. Kancharla, B. Meka, Y.L.N. Murthy, Med. Chem. Res. 25 (2016) 2906–2915. doi: 10.1007/s00044-016-1709-5
Scheme 1 (A) Typical skeletons of classical lignans & representative bioactivemolecules. (B) Biosynthesis pathway & retrosynthetic analysis of dehydrodibenzylbutyrolactone lignans. (C) Desired approach: modular biomimetic synthesis of lignans.
Figure 1 Optimization of the reaction conditions. Standard conditions: 1a (0.2 mmol), 2a (0.6 mmol), Pd(TFA)2 (10 mol%), DPPF (15 mol%), BQ (200 mol%), THF (2 mL), 12 h. The ratio of products were determined by GC–MS with n-dodecane as the internal standard except for isolated yield under standard conditions. TPFPP = tris(pentafluorophenyl)phosphine, BQ = 1, 4-benzoquinone, 2-Me-BQ = p-toluquinone, TBHP = tert‑butyl hydroperoxide, 2-MeTHF = 2-methyltetrahydrofuran.
Scheme 2 The substrate scope of modular biomimetic synthesis of lignans. Isolated yields on 0.2 mmol scale. The dr ratio of 3bs and 3bt were determined by the 1H NMR.
Scheme 3 Molecular hybridization of lignans with drugs. The dr ratio were determined by the 1H NMR.
Scheme 4 Collective synthesis of lignan skeletons. Conditions: (a) NiCl2, NaBH4, THF-MeOH; (b) Mg, MeOH, NH4Cl; (c) DIBALH, toluene; (d) TiCl4, Et3SiH, DCM; (e) LiAlH4, THF; (f) l-selectride, THF; (g) Fe(ClO4)3·6H2O, CF3CO2H; (h) DDQ, CF3CO2H; (i) 3, 4-(methylenedioxy)phenylboronic acid, TEMPO, KF, Pd(OAc)2, propionic acid. See Supporting information for detailed conditions.
扫一扫看文章
扫一扫关注我们
DownLoad:
下载:
下载: